Secrets of succulence 1
Jamie Males* 2
Department of Plant Sciences, University of Cambridge 3
Downing Street, Cambridge, CB2 3EA, UK 4 *Correspondence: 5 [email protected] 6 +44(0)7792992635 7 Date of submission: 22/11/2016 8 Number of tables/figures: 9 Word count: 10 11 Abstract 12
Succulent plants are iconic components of the florae of many terrestrial ecosystems, but 13
despite having prompted fascination and investigation for centuries, they still harbour many 14
secrets in terms of physiological function and evolution. Tackling these mysteries is 15
important, as this will not only provide insights into the dynamics and details of the 16
convergent evolution of a major adaptive syndrome, but also inform efforts to conserve 17
endangered biodiversity and utilise the unique physiological characteristics of succulents for 18
biofuel and biomass production. Here I review advances in the phylogeny and organismal 19
biology of succulent plants, and discuss how insights from recent work in the wider fields of 20
plant hydraulics and photosynthetic physiology may relate to succulents. The potential for 21
the exploration of mechanistic relationships between anatomical structure and physiological 22
function to improve our understanding of the constraints that have shaped the evolution of 23
succulence is highlighted. Finally, attention is drawn to how new methodologies and 24
technologies provide exciting opportunities to address the wide range of outstanding 25
questions in succulent plant biology. 26
Introduction 28
Succulent plants have been the subject of fascination for centuries, but their relevance as 29
masters of water management has perhaps never been greater than now, as with 30
accelerating global change and pressure on natural and agricultural systems urgently 31
demandings urgent insights into the mechanisms of drought-resistance. Understanding the 32
full story of succulent plant biology requires answers to a series of superficially seemingly 33
straightforward, but actually rather challenging, questions. What exactly is succulence? 34
Which plants have evolved succulence, and under what conditions? What selective 35
advantages can succulence confer? What can succulence do for us? In this review, I discuss 36
recent advances towards answering these overarching questions, with a particular emphasis 37
on water relations, and identify a path to take us forwards in the twofold quest both to 38
understand succulent plants and to utilise that understanding in applied contexts. Although 39
some aspects of the distinctive biology of succulent halophytes are briefly discussed, the 40
focus is on classical drought-avoidance succulents (sensu Ogburn and Edwards, 2010). 41
42
What’s in a name? Measuring succulence in its many forms 43
Succulence is a phenomenon that has long eluded a decisive consensus definition. 44
Traditionally, succulent plants have been treated as a distinct functional group within the 45
plant kingdom. The boundaries defining membership of that group have fluctuated, and 46
quite different terms have been used to define them. Few but the most practical of 47
taxonomists would use the definition of the 18th-century botanist Richard Bradley, who 48
identified succulents as those species which are ‘not capable of an Hortus-siccus’ (i.e. could 49
not be prepared as herbarium specimens because of their juiciness; Bradley, 1716-1727). 50
The morphological Gestalt of succulent plants, as described by Ogburn and Edwards (2010), 51
remains a useful concept because of its familiarity, and the binary discrimination between 52
succulents and non-succulents is often adequate in simple functional type classification 53
schemes. However, it is problematic for in terms of the identification of thresholds. What 54
particular combination of trait values are sufficient to make a plant ‘succulent’? Do different 55
succulent plants even conform to a single set of criteria? As will be discussed, superficially 56
equivalent succulent morphologies may be underpinned by strongly contrasting internal 57
anatomy. Indeed, while succulence is manifested fundamentally at the cellular level, this 58
need not translate to morphological succulence. Ogburn and Edwards (2010) give the 59
example of the bromeliad Tillandsia usneoides (L.) L. (Spanish moss), which displays strongly 60
succulent cells, with important consequences for the species’ physiological ecology (Kluge et 61
al., 1973), even though the leaves appear (and the whole plant) are highly morphologically 62
reduced. 63
The enigmatic nature of succulence is perhaps to be expected of any syndrome emerging 64
from variation in quantitative traits (Ogburn and Edwards, 2010). Eggli and Nyffeler (2009) 65
have provided one of the most complete definitions of succulence as the ‘storage of 66
utilizable water in living tissues in one or several plant parts in such a way as to allow the 67
plant to be temporarily independent from external water supply but to retain at least some 68
physiological activity’. According to this definition, succulents must be able to use some of 69
the water they have stored through the regulation of processes in living cells. High 70
apoplastic water content is therefore not sufficient qualification. Nor do succulents enter a 71
state of metabolic inactivity during periods of reduced water availability, as is the case with 72
resurrection plants (Farrant and Moore, 2011). These characteristics sum to make succulent 73
plants classical examples of drought-avoiders (Eggli and Nyffeler, 2009; Ogburn and 74
Edwards, 2010). 75
Some of the many proposed metrics for succulence have been discussed by Von Willert et 76
al. (1990) and Ogburn and Edwards (2010, 20123). While some are based simply on water 77
content, others take into account tissue structure and chemical composition or other 78
anatomical parameters. The easily-quantifiable saturated water content (SWC; Ogburn and 79
Edwards, 2012), which is the ratio of water mass at full hydration to dry tissue mass, is 80
gaining traction in comparative studies. As with all such indices, of prime importance is the 81
principle of comparability. Is what makes one species succulent the same as what makes 82
another species succulent? To answer this, one must consider some of the structural 83
diversity that exists among succulent plants. 84
85
Anatomical and morphological diversity 86
Succulence can occur in any vegetative organ. Although leaf- and stem-succulence are most 87
familiar, water storage may also occur in roots, the bulbs or tubers of geophytes, orchid 88
pseudobulbs, and the parenchymatous rays of pachycaul trees (Eggli and Nyffeler, 2009, 89
2010). Although most physiological research has focussed on stem- and leaf-succulence, 90
Hearn et al. (2013) have shown a high degree of phylogenetic coordination between origins 91
of aboveground and belowground succulence across the eudicots. This suggests that 92
evolutionary transitions in the organ-specificity of succulence can occur quite readily, which 93
in turn points to a common developmental basis of succulence in different plant parts. 94
Within specific organs, succulence can arise from different tissues. For instance, in succulent 95
Peperomia Ruiz & Pav. (Piperaceae) it is primarily the epidermal layers that are involved in 96
water storage (Kaul, 1977), whereas in succulent bromeliads it is the hypodermal layer that 97
has been co-opted for this function (Tomlinson, 1969). 98
Among species with photosynthetic succulent stems and leaves, two main types of 99
anatomical arrangement prevail. Ihlenfeldt (1985) termed these Allzellsukkulenz (‘all-cell 100
succulence’) and Speichersukkulenz (‘storage succulence’). The former term describes the 101
situation where water is stored in enlarged photosynthetic cells, whereas the latter 102
describes a division of labour between photosynthetic tissues and specialised water storage 103
tissues (hydrenchyma). Fig. 1 illustrates some arrangements of chlorenchyma and 104
hydrenchyma that occur in different leaf-succulent lineages. In Fig. 1a, a typical all-cell 105
succulent leaf structure is shown, which involves a comparatively homogenous structure 106
throughout the leaf. The arrangement in Fig. 1b, with a central core of hydrenchyma 107
transitioning either gradually or abruptly into a peripheral rind of chlorenchyma, is typical of 108
many monocot leaf-succulents in the Asparagales (e.g. Aloë spp. and Agave L. spp.). In some 109
succulent groups (e.g. Piperaceae), the reverse arrangement often occurs, with a peripheral 110
layer of hydrenchyma encircling a central core of chlorenchyma. Meanwhile all-cell 111
succulence in the chlorenchyma combined with a well-developed adaxial layer of 112
hydrenchyma is characteristic of many bromeliad species, where the transition between the 113
chlorenchyma and hydrenchyma can be either abrupt (as in Fig. 1c) or more gradual. While 114
it seems likely that this extensive structural variation could account for ecophysiological 115
divergences among leaf-succulents, attempts to definitively draw together interacting 116
structure-function relationships in three-dimensional tissues are only now becoming 117
possible through the emergence of new visualisation and modelling methodologies 118
(Brodersen and Roddy, 2016; Ho et al., 2016). 119
[FIGURE 1] 120
However, Ihlenfeldt (1985) made several suggestions as to the functional significance of the 121
distinction between all-cell succulence and storage succulence in , discussed here in the 122
context of leaves. First, all-cell succulence should be self-limiting with respect to organ size. 123
A larger leaf will hold more water and have a lower surface area-to-volume ratio (SA:V), 124
reducing the ratio of transpiration to hydraulic capacitance. However, thicker tissues impose 125
stronger constraints on the diffusion of CO2 from stomata to chloroplasts, such that 126
assimilation in the centre of the leaf may be inefficient (Maxwell et al., 1997). Perhaps for 127
this reason, all-cell succulence generally occurs in species with small, non-spheroid leaves 128
with a higher SA:V. This has important implications for leaf economics, thermal physiology 129
and light relations. Ihlenfeldt (1985) also remarked that all-cell succulents can only lose a 130
limited amount of water content before experiencing physiological dysfunction, since water 131
loss must necessarily occur from photosynthetically-active cells. 132
Meanwhile, Ripley et al. (2013) have demonstrated that storage-type anatomy can be 133
associated with relatively high chlorenchyma CO2 conductance (gm). The segregation of 134
photosynthetic and water storage functions thus allows gm and photosynthetic capacity to 135
be decoupled from total leaf water content. However, despite this advantage, storage 136
succulence requires investment in mechanical adaptations at a considerable carbon cost 137
(Von Willert et al., 1990), often including a rigid epidermal-hypodermal complex, which 138
Ihlenfeldt (1985) described as a supportive ‘exoskeleton’. Although comparative 139
physiological data are limited, all-cell succulents are generally thought to occupy a position 140
closer to the ecologically opportunistic ‘live fast, die young’ end of the leaf economics 141
spectrum when compared with the more conservative and less flexible storage succulents 142
(Ihlenfeldt, 1985; Von Willert et al., 1990). Many succulents display a combination of all-cell 143
and storage succulence, including members of the Bromelioideae (Bromeliaceae; Tomlinson, 144
1969). 145
Gross morphology, particularly SA:V, is an important determinant of functional succulence. 146
Working with columnar cacti, Williams et al. (2014) elucidated the quantitative links 147
between species-specific stem SA:V, which is constrained by a trade-off between area-based 148
water loss and water storage capacity, and bioclimatic relations. Insights from stable isotope 149
analyses have recently added a third dimension to the picture for cacti: photosynthetic 150
capacity, which is constrained by diffusive and optical trade-offs to evolve in coordination 151
with morphology and climate envelope (Hultine et al., 2016). 152
Leaf temperature is one of the many ecophysiological variables with which succulence 153
interacts through morphology (Nobel, 1988; Von Willert et al., 1992). Both modelling (e.g. 154
Leigh et al., 2012) and empirical work (Larcher et al., 2010; Monteiro et al., 2016) have 155
highlighted the importance of leaf thickness and density for maintaining sub-critical leaf 156
temperature under strong environmental forcing. Additionally, temperature gradients 157
within leaves have recently been implicated in the magnitude of vapour-phase fluxes of 158
water from evaporative sites to the stomatal pore (Rockwell et al., 2014; Buckley, 2015; 159
Buckley et al., 2017). These gradients are likely to be particularly steep in succulent leaves 160
with high thermal capacity. It is possible that some evolutionary origins of succulence may 161
have been promoted byrelated in part to a selective advantage associated with the 162
suppression of the potential for large vapour-phase fluxes. 163
164
Phylogenetic and biogeographic diversity 165
The numerous origins of succulence scattered across the land plant phylogeny are 166
frequently cited as a classic example of morphological (if not anatomical and functional) 167
convergence. While succulence is by no means limited to the angiosperms (it occurs, for 168
instance, in Pyrrosia Mirb. ferns and the gymnosperm Welwitschia Hook.f.), the majority of 169
succulents are flowering plants. Succulents are widely distributed across the angiosperm 170
phylogeny, offering extensive evolutionary replication for investigators (Ogburn and 171
Edwards, 2010). Recently, advances have been made in clarifying phylogenetic relationships 172
within several major succulent lineages, including Aloë L. (Asphodelaceae; Grace et al., 173
2015), Euphorbia L. (Euphorbiaceae; Horn et al., 2012; Peirson et al., 2013; Evans et al., 174
2014; Horn et al., 2014), Opuntia Mill. (Cactaceae; Majure et al., 2012), and Ruschieae 175
(Aizoaceae; Klak et al., 2013). The Portullugo clade (Caryophyllales) developed as a model 176
system by Edwards and colleagues has proved particularly fruitful for testing evolutionary 177
hypotheses (Nyffeler et al., 2008; Ogburn and Edwards, 2009, 2013, 2015). However, there 178
is still tremendous scope for integrated progress in the phylogenetics, morphoanatomy and 179
physiology of such critical taxa as the Crassulaceae, Orchidaceae, Asphodelaceae, 180
Asteraceae, Aizoaceae, Apocynaceae and Bromeliaceae. Improved characterisation of the 181
evolutionary trajectories leading to succulence in different lineages would help us to 182
understand the extent of parallelism in independent origins. 183
[FIGURE 2] 184
Succulents occur in almost all parts of the world, but centres of diversity are readily 185
identifiable (Fig. 2). The deserts and semi-deserts of southwest North America are rich in 186
iconic stem- succulent cacti and leaf- and stem-succulent agaves and Crassulaceae. The 187
forests of the northern Andes host the greatest concentration of succulent epiphytic 188
bromeliads and orchids, although these are widespread throughout the Neotropics and (in 189
the case of the orchids) other tropical regions. Further south in the Andean cordillera is 190
another succulent hotspot reaching from Peru into Bolivia, where cacti and terrestrial 191
bromeliads are particularly profuse. The florae of the Caatinga and Campo Rupestre regions 192
of Brazil include numerous endemic stem-succulent cacti and euphorbs. The highest 193
succulent diversity occurs in southern Africa’s Succulent Karoo, including abundant 194
Aizoaceae, Crassulaceae, caudiciforms and geophytes. In Madagascar caudiciforms are 195
joined by euphorbs and endemic Didiereaceae. Along the North African littoral and on the 196
Macaronesian islands are further radiations of Crassulaceae and Euphorbiaceae, and in the 197
Irano-Turanian floristic region the succulent halophytes of the Chenopodioideae and 198
Zygophyllaceae reach their highest diversity. Other regional florae with notable but less 199
diverse succulent elements include those of Australia and various alpine regions. With the 200
exception of the special case of the northern Andean forests with its diverse epiphyte flora, 201
these hotspots show varying degrees of aridity and seasonality, which are two of the 202
environmental pressures classically associated with succulent growth-forms (Von Willert et 203
al., 1992; Ogburn and Edwards, 2010). 204
There is great disparity in the species richness of succulent clades. The lone succulent grass 205
species, Dregeochloa pumila (Nees) Conert, might be regarded as an evolutionary ‘dead-206
end’ when contrasted with the extensive radiations of other succulent monocot groups like 207
Agave and Aloë, which together comprise over 700 species. The most dramatic succulent
radiations have arisen from what Donoghue and Sanderson (2015) refer to as the 209
‘confluence’ (i.e. co-occurrence) of a ‘synnovation’ and ecological opportunity. 210
‘Synnovation’ denotes an ensemble of adaptive innovations that synergistically displace or 211
broaden a population’s ecological amplitude. Meanwhile, the ecological opportunity is 212
provided by the favourable alignment of environmental factors opening up highly 213
unsaturated niche space to invasion. Recent research has unearthed several examples of 214
this scenario, including Agave, the Aizoaceae, terrestrial Bromeliaceae, Cactaceae and 215
Euphorbiaceae, all of which independently evolved a synnovation complex involving 216
succulence and Crassulacean acid metabolism (CAM). In each case, this synnovation 217
complex was closely linked to exploitation of the large geographical regions of semi-arid 218
climate that arose during the global climatic changes between the late Oligocene and late 219
Miocene (Horn et al., 2014; Good-Avila et al., 2006; Arakaki et al., 2011; Givnish et al., 2014; 220
Hernández-Hernández et al., 2014; Valente et al., 2014). Parallel and contemporaneous 221
selective pressures therefore appear to have been important in shaping the present-day 222
diversity of succulent plants. However, other innovations, including new habits and growth-223
forms (Givnish et al., 2014; Hernández-Hernández et al., 2014; Givnish et al., 2015; 224
Freudenstein and Chase, 2015) and environmental and biotic factors, including forest 225
dynamics (Xiang et al., 2016) and pollinator coevolution (Hernández-Hernández et al., 2014; 226
Givnish et al., 2015; Freudenstein and Chase, 2015), have sometimes been critical. 227
228
Succulence and plant economic relationships 229
Succulence does not represent a single peak on a simple adaptive landscape, because it 230
assumes many primary and secondary functions, ranging from short- to long-term water-231
storage, and from salt accumulation to thermal insulation. Succulence is compatible with 232
occupation of a range of positions along the plant economic spectrum (Reich, 2014), with 233
many storage succulents being slow-growing stress-tolerators, and all-cell succulents being 234
more resource-acquisitive. The diversity of economic strategies displayed by succulents can 235
be expanded even further when drought-deciduous succulents and deciduous leaf-236
succulent geophytes are considered (e.g. Von Willert et al., 1990; Donatz and Eller, 1993; 237
Wiegand et al., 2000). Moreover, the transformative effect of succulence on structure and 238
function is reflected in the way it tends to distort plant economic relationships (Vendramini 239
et al., 2002). For example, the classical correlation between photosynthetic capacity and 240
leaf mass per unit area (LMA; Wright et al., 2004) is notably weaker in leaf-succulents than 241
in other plant groups (Ripley et al., 2013; Grubb et al., 2015). This is because investment in 242
differentiated hydrenchyma introduces an additional source of variation in LMA, but may 243
have comparatively little effect on the photosynthetic capacity of the chlorenchyma. Thus, 244
by rewiring trait networks, origins of succulence can reshape the constraints on functional 245
trait evolution. This important effect could allow new trait combinations to arise and 246
thereby act as a pump for the evolution of ecophysiological diversity. 247
248
Phylogenetic and biogeographic diversity 249
Succulence is not limited to the angiosperms, occurring, for instance, in Pyrrosia Mirb. ferns 250
and the gymnosperm Welwitschia Hook.f. However, the majority of succulents are flowering 251
plants, and they are widely distributed across the angiosperm phylogeny, offering extensive 252
evolutionary replication for investigators (Ogburn and Edwards, 2010). Recently, advances 253
have been made in clarifying phylogenetic relationships within several major succulent 254
lineages, including Aloë L. (Asphodelaceae; Grace et al., 2015), Euphorbia L. (Euphorbiaceae; 255
Horn et al., 2012; Peirson et al., 2013; Evans et al., 2014; Horn et al., 2014), Opuntia Mill. 256
(Cactaceae; Majure et al., 2012), and Ruschieae (Aizoaceae; Klak et al., 2013). The Portullugo 257
clade (Caryophyllales) developed as a model system by Edwards and colleagues has proved 258
particularly fruitful for testing evolutionary hypotheses (Nyffeler et al., 2008; Ogburn and 259
Edwards, 2009, 2013, 2015). However, there is still tremendous scope for integrated 260
progress in the phylogenetics, morphoanatomy and physiology of such critical taxa as the 261
Crassulaceae, Orchidaceae, Asphodelaceae, Asteraceae, Aizoaceae, Apocynaceae and 262
Bromeliaceae. Improved characterisation of the evolutionary trajectories leading to 263
succulence in different lineages would help us to understand the extent of parallelism in 264
independent origins. 265
[FIGURE 1] 266
Succulents occur in almost all parts of the world, but centres of diversity are readily 267
identifiable (Fig. 1). The deserts and semi-deserts of southwest North America are rich in 268
iconic stem- succulent cacti and leaf- and stem-succulent agaves and Crassulaceae. The 269
forests of the northern Andes host the greatest concentration of succulent epiphytic 270
bromeliads and orchids, although these are widespread throughout the Neotropics and (in 271
the case of the orchids) other tropical regions. Further south in the Andean cordillera is 272
another succulent hotspot reaching from Peru into Bolivia, where cacti and terrestrial 273
bromeliads are particularly profuse. The florae of the Caatinga and Campo Rupestre regions 274
of Brazil include numerous endemic stem-succulent cacti and euphorbs. The highest 275
succulent diversity occurs in southern Africa’s Succulent Karoo, including abundant 276
Aizoaceae, Crassulaceae, caudiciforms and geophytes. In Madagascar caudiciforms are 277
joined by euphorbs and endemic Didiereaceae. Along the North African littoral and on the 278
Macaronesian islands are further radiations of Crassulaceae and Euphorbiaceae, and in the 279
Irano-Turanian floristic region the succulent halophytes of the Chenopodioideae and 280
Zygophyllaceae reach their highest diversity. Other regional florae with notable but less 281
diverse succulent elements include those of Australia and various alpine regions. With the 282
exception of the special case of the northern Andean forests with its diverse epiphyte flora, 283
these hotspots show varying degrees of aridity and seasonality, which are two of the 284
environmental pressures classically associated with succulent growth-forms (Von Willert et 285
al., 1992; Ogburn and Edwards, 2010). 286
Among the angiosperms, leaf-succulence is perhaps the most phylogenetically widespread 287
form of succulence at the familial level, with instances of stem- and root-succulence, 288
pachycauly, and succulent tubers or bulbs scattered across the major clades (Nyffeler and 289
Eggli, 2010). However, there is extensive structural variation both between and within 290
families expressing each of these types of succulence. As an example, Fig. 2 illustrates some 291
arrangements of chlorenchyma and hydrenchyma that occur in different leaf storage-292
succulent lineages. The arrangement in Fig. 2a, with a central, sharply-defined core of 293
hydrenchyma, is typical of Aloë spp., whereas a more gradual transition between tissue 294
types is common in Agave spp. (Fig. 2b). A well-developed adaxial layer of hydrenchyma is 295
characteristic of many bromeliad species, where its transition into the chlorenchyma can be 296
either abrupt (Fig. 2c) or gradual (Fig. 2d). In some Piperaceae there is a peripheral layer of 297
hydrenchyma encircling a central core of chlorenchyma (Fig. 2e). While it is intuitive that 298
this extensive structural variation could account for ecophysiological divergences among 299
leaf-succulents, attempts to definitively draw together interacting structure-function 300
relationships in three-dimensional tissues are only now becoming possible through the 301
emergence of new visualisation and modelling methodologies (Brodersen and Roddy, 2016; 302
Ho et al., 2016). 303
[FIGURE 2] 304
There is great disparity in the species richness of succulent clades. The lone succulent grass 305
species, Dregeochloa pumila (Nees) Conert, might be regarded as an evolutionary ‘dead-306
end’ when contrasted with the extensive radiations of other succulent monocot groups like 307
Agave L. and Aloë, which together comprise over 700 species. The most dramatic succulent
308
radiations have arisen from what Donoghue and Sanderson (2015) refer to as the 309
‘confluence’ (i.e. co-occurrence) of a ‘synnovation’ and ecological opportunity. 310
‘Synnovation’ denotes an ensemble of adaptive innovations that synergistically displace or 311
broaden a population’s ecological amplitude. Meanwhile, the ecological opportunity is 312
provided by the favourable alignment of environmental factors opening up highly 313
unsaturated niche space to invasion. Recent research has unearthed several examples of 314
this scenario, including Agave, the Aizoaceae, terrestrial Bromeliaceae, Cactaceae and 315
Euphorbiaceae, all of which independently evolved a synnovation complex involving 316
succulence and Crassulacean acid metabolism (CAM). In each case, this synnovation 317
complex was closely linked to exploitation of the large geographical regions of semi-arid 318
climate that arose during the global climatic changes between the late Oligocene and late 319
Miocene (Horn et al., 2014; Good-Avila et al., 2006; Arakaki et al., 2011; Givnish et al., 2014; 320
Hernández-Hernández et al., 2014; Valente et al., 2014). Parallel and contemporaneous 321
selective pressures therefore appear to have been important in shaping the present-day 322
diversity of succulent plants. However, other innovations, including new habits and growth-323
forms (Givnish et al., 2014; Hernández-Hernández et al., 2014; Givnish et al., 2015; 324
Freudenstein and Chase, 2015) and environmental and biotic factors, including forest 325
dynamics (Xiang et al., 2016) and pollinator coevolution (Hernández-Hernández et al., 2014; 326
Givnish et al., 2015; Freudenstein and Chase, 2015), have sometimes been critical. 327
328
Succulence and plant economic relationships 329
The numerous origins of succulence scattered across the angiosperm phylogeny are 330
frequently cited as a classic example of convergent evolution. However, succulence does not 331
represent a single peak on a simple adaptive landscape, because it assumes many primary 332
and secondary functions, ranging from short- to long-term water-storage, and from salt 333
accumulation to thermal insulation. Furthermore, the transformative effect of succulence 334
on structure and function is reflected in the way it tends to distort plant economic 335
relationships (Vendramini et al., 2002). (Vendramini et al., 2002). For example, the classical 336
correlation between photosynthetic capacity and leaf mass per unit area (LMA) is notably 337
weaker in leaf-succulents than in other plant groups (Ripley et al., 2013; Grubb et al., 2015). 338
This is because investment in differentiated hydrenchyma introduces an additional source of 339
variation in LMA, but may have comparatively little effect on the photosynthetic capacity of 340
the chlorenchyma. Thus, by rewiring trait networks, origins of succulence can reshape the 341
constraints on functional trait evolution. The proximity of any given succulent phenotype to 342
the nearest adaptive peak is also highly dependent on spatiotemporal context. How this 343
rugged, shifting fitness landscape is likely to be remodelled by ongoing environmental 344
change should be prioritised. 345
346
Selection for succulence 347
High degrees of succulence have traditionally been associated with regions of low, seasonal 348
rainfall, and many succulent plants conform to the stereotype of a large, slow-growing 349
perennial in a semi-arid habitat, including most succulent Cactaceae and Euphorbiaceae. 350
However, Ogburn and Edwards (2015) recently demonstrated that in the Montiaceae there 351
is no relationship between succulence (quantified as SWC) and precipitation seasonality, 352
although SWC did correlate negatively with mean annual precipitation. This highlights the 353
need to move on from limiting generalisations. The achievement of a comprehensive 354
understanding of the relationship between succulence and water availability regimes 355
depends on nuanced consideration of the integrative biology of individual taxa on a case-by-356
case basis. 357
One important observation discussed recently is that the climatic conditions in regions in 358
which morphologically analogous succulent taxa occur are not as comparable as previously 359
assumed (Alvarado-Cárdenas et al., 2013; Holtum et al., 2016; see also Moncrieff et al., 360
2015). By definition, the florae of hotspots of succulent diversity are composed of a high 361
proportion of endemics. While there is a long tradition of analysing endemicity in the 362
context of phylogenetic identity, emphasis should now be placed on establishing the 363
relationships between endemicity, form and function. Even where characteristic taxa of 364
different geographical regions appear superficially analogous in morphology, they may 365
diverge in physiological function thanks to subtle dissimilarities in anatomy. 366
Water limitation is not only a function of macroclimatic variation; the connection between 367
the epiphytic habit and adaptations for conservative water use has long been 368
acknowledged. Epiphytism is characteristic of several major radiations of vascular plants, 369
including polypod ferns, epidendroid orchids, bromeliads, gesneriads, many of which are 370
succulent (Nyffeler and Eggli, 2010). Although not all epiphytes show pronounced 371
succulence, it is notable that very low degrees of succulence are most common in epiphytes 372
that have evolved phytotelmata as external hydraulic capacitors (e.g. the tank bromeliads; 373
Males, 2016). Selection for succulence is maintained even among epiphytes inhabiting 374
montane cloud forests (e.g. Gotsch et al., 2015) and temperate rainforests (e.g. Godoy and 375
Gianoli, 2013), underlining the difficulties of water acquisition in the absence of soil rooting. 376
Succulent plants are also well represented in alpine environments. Temperate examples 377
including species in genera such as Sedum L. and Sempervivum L. in the Crassulaceae 378
(Codignola et al., 1990), while tropical examples include giant rosette species in Espeletia 379
Mutis ex Bonpl. In Humb. & Bonpl. (Asteraceae) and Lobelia L. (Campanulaceae; Carlquist, 380
1994). Plants growing at high elevations experience numerous intense environmental 381
pressures, often including water limitation, but also extreme temperatures and ultraviolet 382
(UV) exposure. Succulence may be beneficial with respect to the latter two pressures as well 383
as its more obvious role in plant water economy. The high thermal capacity of massively 384
succulent leaves can effectively uncouple them from low atmospheric temperatures at 385
night, helping to protect cold-sensitive critical tissues (i.e. the shoot apical meristem; Nobel, 386
1988). Morphological adaptation including pubescence can also modulated night-time leaf 387
temperatures (e.g. Keeley and Keeley, 1989). Many succulents inhabiting locations where 388
temperatures drop below 0°C also display structural or biochemical adaptations to avoid 389
freezing injury (e.g. Nobel and De La Barrera, 2003). High temperature tolerance is also 390
common in alpine succulents (e.g. Larcher et al., 2010). Alongside the epidermal 391
specialisation to improve UV reflectance (Mulroy, 1979) and high investment in antioxidant 392
phenolics (Bachereau et al., 1998) that are often observed in alpine succulents, it is possible 393
that species with peripheral hydrenchyma could benefit from increased UV reflectance by 394
this tissue. 395
Halophytes are often described as using succulence to cope with physiological drought 396
rather than the physical water shortage faced by drought-avoidance succulents. Succulent 397
halophytes are epitomised by species of the Chenopodioideae and Salicornioideae 398
(Amaranthaceae; Flowers and Colmer, 2015). However, halophytes are very different in 399
their water-use strategies and their relationship with succulence. Ogburn and Edwards 400
(2010) suggested that succulence in halophytes is primarily a by-product of ionic 401
accumulation in enlarged vacuoles and does not provide capacitance. Halophytic succulence 402
is therefore an almost completely distinct phenomenon, and there are very few examples of 403
angiosperm lineages that display both halophytic and drought-avoidance succulence 404
(Ogburn and Edwards, 2010). 405
406
The physiology of succulent water use 407
The physiology of water use in succulent plants varies more widely than is often suggested. 408
While transpiration rates are strongly restricted in xerophytic drought-avoidance succulents, 409
it has long been recognised that they can be relatively high in succulent halophytes (Delf, 410
1911, 1912). Among drought-avoidance succulents, two contrasting strategies can be 411
identified in terms of the seasonal dynamics of stored water use. These two strategies are 412
closely connected to life-history. 413
In small annual succulents, including many Aizoaceae, succulent organs represent single-use 414
water stores that can extend the growing season into the portion of the year defined by less 415
favourable climatic conditions, and depletion of the store coincides with seed production 416
and senescence (Ogburn and Edwards, 2015). A very different type of hydraulic behaviour is 417
observed in storage succulents. These plants display a distinctive water-use strategy 418
involving translocation of water from succulent storage tissue to chlorenchyma during 419
seasonal drought (e.g. Nobel, 2006), buffering chlorenchyma water potential, followed by 420
refilling of hydrenchyma during seasonal precipitation events. The rehydration process can 421
occur quickly (Scalisi et al., 2016), and involves coordinated responses of root and shoot 422
tissues (North et al., 2004; Griffiths, 2013). In some Agave species, a network of fine, short-423
lived ‘rain roots’ rapidly develops, increasing total root length by 47% in Agave deserti 424
Engelm. (Jordan and Nobel, 1984). In the shoot, aquaporins are also important in 425
maximising the conductance of the pathway between the vasculature and the storage 426
tissues (North et al., 2004). Stomatal aperture also increases, generating a stronger 427
transpirational pull that may help to draw water through the plant towards storage tissues 428
as well as towards the stomata (Nobel, 1988). 429
[FIGURE 34] 430
A range of anatomical and biochemical factors are likely to influence the capacity for 431
efficient recharge by modifying the overall hydraulic resistance of the root-capacitor 432
pathway and the partitioning of relative resistances between xylary and extra-xylary 433
compartments (Fig. 34). The overall efficiency of the process should be maximised by 434
coordinated evolutionary changes in xylem properties and processes such as the ionic effect 435
(Zwieniecki et al., 2001), but also in the aquaporin profiles and anatomy of both the root 436
and shoot. Interveinal distance is generally positively correlated with succulence due to 437
developmental constraints imposing a limitation on hydraulic connectivity in many 438
succulents (Ogburn and Edwards, 2013). However, in some succulent lineages, there has 439
been convergent evolution of ‘three-dimensional’ arrangements of vascularisation, in which 440
multiple layers of vascular bundles permeate the mesophyll. This allow hydraulic 441
homogeneity to be preserved in more succulent leaves by maintaining a low path length for 442
water transport between veins and mesophyll cells (Ogburn and Edwards, 2013; Melo-de-443
Pinna et al., 2016). This innovation has been linked with elevated degrees of succulence and 444
rates of diversification in the Portulacineae and Molluginaceae (Ogburn and Edwards, 2013). 445
Similar phenomena have been described in the vasculature of succulent stems in other taxa 446
(Mauseth, 1993; Carlquist, 2001; Hearn, 2009). Further empirical and modelling work is 447
required to tease apart the anatomical and biochemical traits that determine the efficiency 448
of recharge and of subsequent water retention. 449
An important aspect of the vascular biology of succulents that warrants further attention in 450
the context of succulence is the organographic disposition of vessel elements in the xylem. 451
Carlquist (2009, 2012) has pointed out that in monocots, where root and shoot xylem are 452
discontinuous, succulence is generally accompanied by the restriction of vessel elements to 453
the root and the presence of tracheids or at most low-diameter vessels in the shoot. This 454
arrangement is hypothesised to facilitate the rapid uptake of transiently-available water by 455
the root system but low hydraulic conductance and water loss from aerial organs. 456
Preferential loss of conductance in roots and leaves, due either to cavitation or extra-xylary 457
effects, could serve to hydraulically isolate the stem from declining soil water potential, 458
reducing the chances of catastrophic hydraulic dysfunction during extreme drought (Linton 459
and Nobel, 1999, 2001; North et al., 2004). This is an example of hydraulic segmentation. 460
There is currently intense interest in this phenomenon in the plant hydraulics community 461
(Pivaroff et al., 2014; Bouche et al., 2016; Hochberg et al., 2016; Johnson et al., 2016; Savi et 462
al., 2016; Wolfe et al., 2016; Zhu et al., 2016). Many stem succulents are drought-deciduous 463
(e.g. Adenium spp.), but how the prevalence of this phenomenon is modulated by the 464
differential distribution of succulence between plant organs remains unclear. It might be 465
expected to be influenced by the construction costs of hydrenchyma and by shoot 466
architecture. 467
Within angiosperm leaves, extra-xylary resistance is often equal to or in excess of xylem 468
resistance (Cochard et al., 2004), depending on environmental conditions (Ocheltree et al., 469
2013). This is probably especially true of for succulents, where the extra-xylary hydraulic 470
pathway is long and tortuous. New models have recently been developed to investigate the 471
relationships between extra-xylary leaf anatomy and hydraulic conductance (Buckley et al., 472
2015, 2017), which should applied to investigate functional differentiation between 473
contrasting succulent anatomies. Furthermore, there is growing evidence that in many plant 474
taxa, the hydraulic conductance of the extra-xylary compartment may be more sensitive to 475
declining water potential than that of the xylem (Scoffoni et al., 2014; Martorell et al., 2015; 476
Trifilò et al., 2016). How these phenomena play out in succulents is not yet known. In many 477
cases, stomatal behaviour is probably sufficiently conservative to minimise the chances of 478
any significant loss of xylem or extra-xylary hydraulic conductance. However, if turgor loss of 479
mesophyll cells is a potential component of extra-xylary hydraulic vulnerability, we might 480
expect this to be particularly important in succulent species. This is because succulents 481
probably lose turgor at relatively modest leaf water potentials due to low osmolarity, as 482
discussed by Martin et al. (2004) in the context of vascular epiphytes. Aquaporins and 483
plasmodesmata are likely to be very important in the hydraulic conductance of succulent 484
organs because of the high degree of cell-cell connectivity associated with succulent 485
anatomy (Steudle et al., 1980; Murphy and Smith, 1998; Buckley et al., 2015). Since 486
aquaporins are the subject of dynamic regulation, they too could play a key role in variable 487
extra-xylary hydraulic conductance, as has been shown in some non-succulent species (e.g. 488
Vitali et al., 2016). 489
The extent to which the hydraulic capacitance of succulent tissues can be dynamically 490
connected to the transpiration stream to buffer transpiration in the face of variable 491
evaporative demand is not clear (Blackman and Brodribb, 2011). Anatomical factors are 492
important determinants of the connectivity between different pools of leaf water, as 493
indicated by rehydration kinetics experiments (Zwieniecki et al., 2007). The physiological 494
processes involved in the remobilisation of stored water in storage succulents warrant 495
further attention, which may require innovation in real-time imaging methodologies. 496
At the distal end of the endogenous transpiration stream, the sensitivity of stomata of 497
succulent plants to environmental stimuli is still poorly documented. It will be interesting to 498
determine whether evidence can be found for modulation of stomatal sensitivity by other 499
tissue-specific hydraulic conductances and capacitances (Ocheltree et al., 2014), or by 500
stomatal density, size or structural diversity (Franks and Farquhar, 2007; Raven, 2014; 501
McAusland et al., 2016). 502
Recognition of interspecific differentiation in water-use strategies has given rise to the 503
elaboration of hydrological niche theory, which is now well-supported for terrestrial plants 504
(Araya et al., 2011; Silvertown et al., 2015). Succulent plants are sometimes caricatured as 505
somewhat monolithic in terms of their water-use strategies, but this is far from accurate. 506
Some terrestrial succulents, including Aizoaceae from coastal southern African deserts are 507
dependent on occult precipitation rather than rainfall (Matimati et al., 2013). Many 508
succulent epiphytes use specialised structures to harvest atmospheric moisture (Reyes-509
García et al., 2008; Zotz and Winkler, 2013). The morphological and physiological variety 510
among co-occurring terrestrial succulents has been shown to support hydrological 511
partitioning (February et al., 2013), and recent modelling efforts have demonstrated how 512
succulent drought-avoidance strategies can coexist with drought-tolerance strategies under 513
water-limited conditions (Manzoni et al., 2014). 514
515
Evolutionary developmental biology of succulence 516
The mechanistic basis of the evolution of succulence remains a puzzle. Relatively little work 517
has been undertaken to explore genetic and ontogenetic mechanisms associated with 518
succulence, or how these differ in storage and all-cell succulence. Hypothetical sequences of 519
stages of structural and physiological specialisation during the evolution of storage and all-520
cell succulence are outlined in Fig. 43. These evolutionary pathways remain largely 521
unexplored, and invite many intriguing questions. For example, are the evolutionary paths 522
to these two types of succulence rigidly parallel from an early stage, or is it possible to 523
‘jump’ from one to the other? Well-resolved phylogenies of key clades are needed to 524
explore these issues, but also better characterisation of the structural detail and selective 525
advantages of different anatomies. At present, we can begin to speculate about how some 526
of the evolutionary changes might have come about. 527
[FIGURE 43] 528
Although apoplastic water and mucilage can make an important quantitative contribution to 529
succulence (Nobel et al., 1992; Ogburn and Edwards, 2009), the largest reservoir of water 530
and that which is under the tightest physiological control resides within living cells (Ogburn 531
and Edwards, 2010). Cell size is therefore an important determinant of succulence. Many 532
factors influence cell size (Marshall et al., 2012), among which is nuclear genome size 533
(Beaulieu et al., 2008). Available data are currently too limited to test for a link between 534
genome size and succulence in a phylogenetically-structured manner. As an alternative to 535
genome size, ploidy level can vary. Polyploidy has been documented in many succulent 536
lineages, but there has been no attempt at systematic review to identify correlations with 537
succulence. Ploidy can also vary within the body of the plant, a phenomenon known as 538
endopolyploidy or endoreduplication. De Rocher (1990) identified a role for endopolyploidy 539
in the development of succulence in Mesembryanthemum crystallinum L., and similar 540
observations have been made in other succulent species (Braun and Winkelmann, 2016). 541
Mishiba and Mii (2000) found higher levels of endopolyploidy in the large hydrenchyma cells 542
of Portulaca grandiflora Hook. than in smaller chlorenchyma and bundle sheath cells. The 543
genetic and developmental determinants of endopolyploidy are not well characterised, but 544
some key regulators have been identified in Arabidopsis, including the STRUWWELPETER 545
(SWP) gene (Autran et al., 2002). 546
Cell size is also affected by the macromolecular content of the cytoplasm, which is 547
controlled by translational regulators such as TARGET OF RAPAMYCIN (TOR) and ErbB-3 548
BINDING PROTEIN1. Overexpression of these factors in Arabidopsis leads to increases in 549
cytoplasmic protein content and cell size (Horváth et al., 2006; Deprost et al., 2007). The 550
vacuolar contribution to cell volume is also important, representing over 90% of the cell 551
volume in succulents (Gibson, 1982; Von Willert et al., 1992). Increased vacuolar ATPase 552
activity is associated with larger cells in Arabidopsis (Ferjani et al., 2013), but otherwise little 553
is known regarding factors controlling vacuole size. 554
For a protoplast to increase in volume, the cell wall must also expand. Overexpression of 555
EXPANSIN10 in Arabidopsis causes an increase in cell size (Cho and Cosgrove, 2000), and 556
Han et al. (2013) have shown that expression of a poplar xyloglucan 557
endotransglucosylase/hydrolase (XTH) transgene in tobacco plants could lead to increased 558
leaf-succulence. Cell wall elasticity is also closely related to capacitance, since it is through 559
changes in cell volume that water is brought in and out of symplastic storage. Cell-cell 560
hydraulic continuity must also be maintained, and the means of achieving this with the 561
greatest potential for dynamic control is to increase the abundance and activity of plasma 562
membrane aquaporins. Qi et al. (2009) demonstrated the importance of an increase in 563
aquaporin activity in the induction of succulence in Suaeda maritima (L.) Dumort, while 564
Vitali et al. (2016) have shown that aquaporins can be involved in the determination of 565
hydraulic capacitance in grapevine. 566
A final consideration in relation to cell size is the phenomenon of compensation. If a 567
mutation causes a decline in cell number, cell size tends to increase in proportion (Hisanaga 568
et al., 2015). The underlying mechanisms of compensation are as yet unknown, as is how 569
the process relates to the evolution of succulence. If maximal succulence is achieved 570
through increases in both cell number and cell size, does this require a loss of function in the 571
machinery of compensation? So far no studies have addressed these issues. 572
An exciting opportunity in succulent evolutionary development lies in naturally-occurring 573
inducible succulence. This phenomenon is known in a range of taxa, notably certain 574
halophytes on exposure to high concentrations of NaCl (Jennings, 1976; Tiku, 1976; Sui et 575
al., 2010). Physiological drought of this kind has been shown to lead to leaf-succulence 576
through endoreduplication in Lobularia maritima (Brassicaceae; Capesius and Loeben, 577
1983). Succulence is also sometimes induced in response to nutrient deficiencies (Baker et 578
al., 1956; Sharma and Ramchandra, 1989; Sharma et al., 1995), while photoperiod regulates 579
the expression of succulence in some Crassulaceae species (Von Denffer, 1941). 580
Transcriptomic and proteomic comparisons of pre- and post-induction tissues from relevant 581
species could prove extremely illuminating. Indeed, increasing numbers of published 582
genomes and transcriptomes of succulent plants could provide an opportunity for 583
comparative analyses across taxa (Gross et al., 2013; Cai et al., 2015; Ming et al., 2015; 584
Hartwell et al., 2016). Intraspecific variation and phenotypic plasticity in succulence is still 585
little-studied, but, due to the complex cost-benefit scenario inherent in the integration of 586
succulence in leaf structure and function, is probably prevalent and ecologically significant. 587
Chiang et al. (2013) recently showed that in the epiphytic fern Pyrrosia lanceolata (L.) Farw., 588
investment in hydrenchyma was strongly influenced by local climatic conditions. 589
590
The complex relationship between succulence and CAM 591
In any discussion of succulence, there is a photosynthetic elephant in the room: 592
Crassulacean acid metabolism (CAM). CAM involves nocturnal stomatal opening and initial 593
fixation of CO2 by phospho-enol-pyruvate carboxylase (PEPC), generating four-carbon 594
organic acids which accumulate in mesophyll cell vacuoles through the course of the night 595
(Osmond, 1978). After dawn, PEPC activity ceases, stomata close, and the stored organic 596
acids are remobilised and decarboxylated to provide extremely high levels of CO2 for 597
RuBisCO-mediated refixation during the light period. Nocturnal stomatal opening enhances 598
water-use efficiency (WUE) since the leaf-air vapour pressure deficit is generally lower at 599
night, and CAM is therefore classically associated with the same environmental pressures as 600
drought-avoidance storage succulence (Osmond, 1978; Lüttge, 2004). Indeed, because of 601
the requirement for large, highly vacuolate mesophyll cells for organic acid storage in CAM, 602
some degree of succulence is required for CAM to be efficient (Zambrano et al., 2014). The 603
efficiency of CAM is also improved in densely-packed, thick tissues, partly due to reductions 604
in leakiness between decarboxylation of organic acids and refixation by RuBisCO (Maxwell et 605
al., 1997; Nelson et al., 2005; Nelson and Sage, 2008). Heyduk et al. (2016a) recently used a 606
case of C3-CAM hybridisation in Yucca (Asparagaceae) to provide microevolutionary insights 607
into the coupling of succulence and CAM. Most origins of succulence have accompanied 608
transitions from C3 to CAM photosynthesis (Ogburn and Edwards, 2010), although it is 609
generally unclear which trait has evolved first, partly because of a paucity of accurate 610
phylogenetic information (Hancock and Edwards, 2014). However, Heyduk et al. (2016b) 611
have recently demonstrated that succulent anatomy predates CAM in the Agavoideae 612
(Asparagaceae), an important radiation of monocot CAM-succulents. Key to further progress 613
in understanding the coordinated evolution of succulence and CAM is the recognition that 614
CAM is a complex syndrome with a continuous scale of functionality rather than a simple 615
binary trait (Silvera et al., 2010; Winter et al., 2015). 616
Many questions surrounding the wider physiological significance of CAM biochemical 617
rhythms in succulents still need to be comprehensively answered. For instance, it remains 618
unclear how the accumulation of osmotically-active compounds during CAM influences 619
internal movements of water in succulent tissues, or whether they might enhance foliar 620
water uptake (Smith and Lüttge, 1985). Similarly, the complex interactions between acidity 621
levels and other aspects of leaf function in CAM-succulents are still imperfectly understood, 622
despite recent advances (Krause et al., 2016). More fundamentally, gaps in our knowledge 623
of the phylogenetic and geographical distribution of succulence and CAM still hamper 624
efforts to understand their relation to climatic factors (Holtum et al., 2016). 625
While CAM is the dominant photosynthetic syndrome among succulents, photosynthetic 626
innovation among succulent plants extends to other pathways. Not only is classical C4 627
photosynthesis with spatial separation of biochemistry between bundle sheath and 628
mesophyll cells common in succulent halophytes (Sage et al., 2011), but single-cell C4 629
systems operate in some Amaranthaceae (e.g. Bienertia; Jurić et al., 2016), and unique C4-630
CAM species occur in Portulaca (Portulaceae; Christin et al., 2014; D’Andrea et al., 2014). 631
632
Solving the secrets of succulence 633
Our understanding of succulence in plants is far from complete. Major questions relating to 634
physiological function, development and evolution remain to be answered. For centuries, 635
succulents have been regarded as curiosities. Eggli and Nyffeler (2009) refer to them as a 636
Sonderfall- a special case- because of their unique biology. Historically, relatively few 637
succulents have enjoyed commercial or agricultural significance. These include the 638
pineapple (Ananas comosus (L.) Merr.), vanilla orchid (Aloë vera (L.) Burm.f.), and Agave 639
spp. used for the production of tequila, mescal and sisal; other succulents are important 640
ornamentals (e.g. orchids, Kalanchoë spp., succulent geophytes). However, it would be 641
timely now to dispense with connotations of oddness and irrelevance, because there are 642
increasingly many practical reasons to be interested in succulent plants. 643
There is growing interest in the use of succulent CAM plants (e.g. Agave, Opuntia) for 644
bioenergy production (Borland et al., 2009; Davis et al., 2011; Holtum et al., 2011; Owen and 645
Griffiths, 2014; Yang et al., 2014; Owen et al., 2016a,b). However, recent studies using new 646
technologies have demonstrated that we do not yet have a clear understanding of the 647
ecophysiological resilience of these plants. Eddy covariance measurements made on a field 648
of Agave tequilana F.A.C. Weber plants showed that gas exchange was unaffected even 649
when soil water potential dropped below the threshold identified by previous studies on 650
individual plants (Nobel, 1988; Owen et al., 2016a). Productivity models based on 651
unrepresentative published parameter estimates could therefore generate misleading 652
results, and further work is needed to explore the complexities of the ecophysiological 653
tolerances of bioenergy candidates. 654
Research programmes are also underway to engineer CAM into C3 plants for bioenergy and 655
food production (Borland et al., 2015). There are many hurdles to clear on the path to 656
successful engineering of CAM (Borland et al., 2014), including the imposition of succulence 657
to provide sufficient vacuolar storage for malic acid produced during CAM. It is therefore 658
essential to develop systems of reliably inducing functional succulent anatomy, including 659
both increased cell volume, organ volume and cell connectivity. 660
There are also gains to be made from enhanced knowledge of succulent physiology 661
in ecological applications. Succulents make a major contribution to the biomass and 662
diversity of regions such as the Succulent Karoo, but a disproportionately high number of 663
succulent species are already considered endangered (e.g. Goettsch et al., 2015). Better 664
Formatted: Font: 12 pt
Formatted: Normal, No bullets or
understanding of the physiological ecology of succulent plants is critical to predicting how 665
the vegetation of fragile ecosystems will respond to climate change (Midgley and Thuiller, 666
2007; Hoffman et al., 2009; Shiponeni et al., 2011; Munson et al., 2012; Schmiedel et al., 667
2012). While succulent plants show variation in the breadth of their environmental 668
tolerance (Midgley and Thuiller, 2007; Schmiedel et al., 2012), the fitness of any given 669
succulent phenotype is generally highly dependent on bioclimatic context. Research into 670
how the fitness landscapes of different groups of succulent plants are likely to be 671
remodelled by ongoing environmental change should be prioritised. For instance, a 672
combination of empirical and modelling work could be undertaken to explore the sensitivity 673
of long-lived storage succulents to alterations in precipitation regime, taking into account 674
both the direct impacts on water storage and the implications for other plant economic 675
traits including photosynthetic potential. 676
Furthermore, improved knowledge of succulent biology may help us to better understand 677
the basis of the economically costly invasiveness of succulents such as Carpobrotus N.E.Br., 678
Lampranthus N.E.Br., and Opuntia spp. (Campoy et al., 2016; Fenollosa et al., 2016). 679
Formatted: Font: 12 pt
Formatted: Font: 12 pt
680
Concluding remarks and future perspectives 681
Important progress towards understanding the evolutionary physiology of succulence has 682
been made in recent years. We now have a clearer picture of the taxonomic distribution of 683
succulence, the evolutionary trajectory is has taken in certain lineages, and the selective 684
advantages it confers in particular environments. Functional divergence between different 685
Box 1. Outstanding Challenges and Opportunities
1. Reconstruction of evolution of succulence by resolution of phylogenetic relationships within major succulent radiations and non-succulent relatives
Improved computing power for large and complex analyses
Accessible and easy-to-use packages for analysis of trait evolution and species diversification rates (e.g. ‘phytools’ for R, Revell, 2012; ‘diversitree’ for R, FitzJohn, 2012) 2. Identification of fixed and dynamic determinants of pathways of
water movement within succulent leaves
Three-dimensional anatomical microstructure visualisation and functional modelling (Brodersen and Roddy, 2016)
Experimental silencing of aquaporins using miRNAs and amiRNAs (Sade et al., 2014, 2015)
3. Quantification of variation in functional traits relevant to ecophysiological differentiation across wider range of understudied succulent lineages
New rapid screening techniques and indices for in situ characterisation of ecophysiological traits (e.g. Bartlett et al., 2012; Ogburn and Edwards, 2012; De Kauwe et al., 2016)
4. Identification of molecular factors involved in the induction of succulence in facultative succulents
Comparative transcriptomic analysis (cf. CAM induction, Brilhaus et al., 2016)
5. Engineering of succulence into non-succulent plants Genome editing techniques (Belhaj et al., 2015)
succulent anatomies has attracted attention, and this has begun to shed light on links 686
between plant structure and climate relations. A comprehensive portrait of the integrative 687
biology of this large and diverse functional group is a long-term goal that will improve our 688
understanding of plant evolution and support successful exploitation of succulence in 689
applied contexts, and there are many areas to which researchers from different 690
backgrounds can make important contributions (see Box 1, Outstanding Challenges and 691
Opportunities). More robust phylogenies of major succulent lineages and their sister taxa 692
are required to reconstruct the evolutionary origins of succulence in finer detail. New 693
empirical work on structure-function relationships is needed, including studies of the 694
interaction between anatomy and aquaporins in controlling tissue water dynamics. This will 695
help to build better models of succulent water use and make predictions of the responses of 696
succulents to environmental fluctuation in natural and agricultural contexts. The natural 697
diversity of succulents should be exploited through molecular screening methodologies to 698
identify key regulatory factors involved in the induction and development of succulence as a 699
means to facilitating efficient engineering of succulence and CAM. 700
701
Acknowledgements 702
Two anonymous reviewers provided constructive feedback on an earlier version of the 703
manuscript. JM is funded by Natural Environment Research Council award 1359020. 704
References
Alvarado-Cárdenas, L.O., Martínez-Meyer, E., Feria, T.P., Eguiarte, L.E., Hernández, H.M., Midgley, G., Olson, M.E. 2013. To converge or not to converge in environmental space: testing for similar environments between analogous succulent plants of North America and Africa. Annals of Botany 111, 1125-1138.
Arakaki, M., Christin, P.-A., Nyffeler, R., Lendel, A., Eggli, U., Ogburn, R.M., Spriggs, E., Moore, M.J., Edwards, E.J. 2011. Contemporaneous and recent radiations of the world’s major succulent plant lineages. Proceedings of the National Academy of Sciences of the U.S.A. 108, 8379-83.
Araya, Y.N., Silverton, J., Gowing, D.J., McConway, K.J., Linder, H.P., Midgley, G. 2011. A fundamental, eco-hydrological basis for niche segregation in plant communities. New Phytologist 189, 253-258.
Autran, D., Jonak, C., Belcram, K., Beemster, G.T.S., Kronenberger, J., Grandjean, O., Inzé, D., Traas, J. 2002. Cell numbers and leaf development in Arabidopsis: a functional analysis of the STRUWWELPETER gene. The EMBO Journal 21, 5955-6287.
Bachereau, F., Marigo, G., Asta, J. 1998. Effect of solar radiation (UV and visible) at high altitude on CAM-cycling and phenolic compound biosynthesis in Sedum album. Physiologia Plantarum 104, 203-210.
Baker, J.E., Gauch, H.G., Dugger, W.M. 1956. Effect of boron on the water relations of plants. Plant Physiology 31, 89-94.
Bartlett, M.K., Scoffoni, C., Ardy, R., Zhang, Y., Sun, S., Cao, K., Sack, L. 2012. Rapid determination of comparative drought tolerance traits: using an osmometer to predict turgor loss point. Methods in Ecology and Evolution 3, 880-888.
Beaulieu, J.M., Leitch, I.J., Patel, S., Pendharkar, A., Knight, C.A. 2008. Genome size is a strong predictor of cell size and stomatal density in angiosperms. New Phytologist 179, 975-986.
Belhaj, K., Chaparro-Garcia, A., Kamoun, S., Patron, N.J., Nekrasov, V. 2015. Editing plant genomes with CRISPR/Cas9. Current Opinion in Biotechnology 32, 76-84.
Blackman, C.J., Brodribb, T.J. 2011. Two measures of leaf capacitance: insights into the water transport pathway and hydraulic conductance in leaves. Functional Plant Biology 38, 118-126.
Borland, A.M., Griffiths, H., Hartwell, J., Smith, J.A.C. 2009. Exploiting the potential of plants with crassulacean acid metabolism for bioenergy production on marginal lands. Journal of Experimental Botany 60, 2879-2896.
Borland, A.M., Hartwell, J., Weston, D.J., Schlauch, K.A., Tschaplinski, T.J., Tuskan, G.A., Yang, X., Cushman, J.C. 2014. Engineering crassulacean acid metabolism to improve water-use efficiency. Trends in Plant Science 19, 327-338.
Borland, A.M., Wullschleger, S.D., Weston, D.J., Hartwell, J., Tuskan, G.A., Yang, X., Cushman, J.C. 2015. Climate-resilient agroforestry: physiological responses to climate change and engineering of crassulacean acid metabolism (CAM) as a mitigation strategy. Plant, Cell & Environment 38, 1833-1849.
Bouche, P.S., Delzon, S., Choat, B. et al. 2016. Are needles of Pinus pinaster more vulnerable to xylem embolism than branches? New insights from X-ray computer tomography. Plant, Cell & Environment 39, 860-870.
Bradley, R. 1716-1727. Historia plantarum succulentarum. London.
Braun, P., Winkelmann, T. 2016. Flow cytometric analyses of somatic and pollen nuclei in midday flowers (Aizoaceae). Caryologia 69, 303-314.
Brilhaus, D., Bräutigaum, A., Mettler-Altmann, T., Winter, K., Weber, A.P.M. 2016. Reversible burst of transcriptional changes during induction of crassulacean acid metabolism in Talinum triangulare. Plant Physiology 170, 102-122.
Brodersen, C.R., Roddy, A.R. 2016. New frontiers in the three-dimensional visualization of plant structure and function. American Journal of Botany 103, 184-188.
Buckley, T.N. 2015. The contributions of apoplastic, symplastic and gas phase pathways for water transport outside the bundle sheath in leaves. Plant, Cell & Environment 38, 7-22. Buckley, T.N., John, G.P., Scoffoni, C., Sack, L. 2015. How does leaf anatomy influence water transport outside the xylem? Plant Physiology 168, 1616-1635.
Buckley, T.N., John, G.P., Scoffoni, C., Sack, L. 2017. The sites of evaporation within leaves. Plant Physiology doi: pp.01605.2016
Cai, J., Liu, X., Vanneste, K. et al. 2015. The genome sequence of the orchid Phalaeonopsis equestris. Nature Genetics 47, 65-72.
Campoy, J.G., Retuerto, R., Roiloa, S.R. 2016. Resource sharing strategies in ecotypes of the invasive clonal plant Carpobrotus edulis: specialization for abundance or scarcity of
resources. Journal of Plant Ecology doi: 10.1093/jpe/rtw073
Capesius, I., Loeben, S. 1983. Changes of nuclear DNA composition after induction of succulence in Lobularia maritima. Zeitschrift für Pflanzenphysiologie 110, 259-266. Carlquist, S.J. 1994. Anatomy of tropical alpine plants. In: Rundel, P.W., Smith, A.P., Meinzer, F.C. (eds.) Tropical alpine environments: plant form and function. Cambridge: Cambridge University Press, 111-128.
Carlquist, S.J. 2001. Comparative wood anatomy: Systematic, ecological, and evolutionary aspects of dicotyledon wood (2nd edn). Berlin: Springer-Verlag.
Carlquist, S.J. 2009. Xylem heterochrony: an unappreciated key to angiosperm origin and diversifications. Botanical Journal of the Linnean Society 161, 26-65.
Carlquist, S.J. 2012. Monocot xylem revisited: new information, new paradigms. Botanical Review 78, 87-153.
Chiang, J.-M., Lin, T.-C., Luo, Y.-C., Chang, C.-T., Cheng, J.-Y., Martin, C.E. 2013. Relationships among rainfall, leaf hydrenchyma, and Crassulacean acid metabolism in Pyrrosia lanceolata (L.) Fraw. (Polypodiaceae) in central Taiwan. Flora 208, 343-350. Cho, H.T., Cosgrove, D.J. 2000. Altered expression of expansin modulates leaf growth and pedicel abscission in Arabidopsis thaliana. Proceedings of the National Academy of Sciences of the U.S.A. 97, 9783-9788.
Christin, P.A., Arakaki, M., Osborne, C.P., Bräutigam, A., Sage, R.F., Hibberd, J.M., Kelly, S., Covshoff, S., Wong, G.K.-S., Hancock, L., Edwards, E.J. 2014. Shared origins of a key enzyme during the evolution of C4 and CAM metabolism. Journal of Experimental Botany 65, 3609-3621.
